BAGGING REGRESI LOGISTIK ORDINAL PADA STATUS GIZI BALITA

Abstract

World Health Organization-National Centre for Health Statistic (WHO-NCHS) is standart nutritional status used in Indonesia, it based on Kartu Menuju Sehat (KMS). These Indices can be expressed in terms of Z-score based Weight-for-Age. This Indices need comparison considering the fact which cause nutritional status not only Weight-for-Age. The aim from this research to obtain bagging ordinal logistics regression for WHO-NCHS nutritional status and new nutritional status. A new nutritional status expressed in terms of cluster, while classification function expressed from logit model of ordinal logistics regression. The result for new nutritional status bagging obtained at 60 bootstrap replicated that is 76.345%, this model can decrease misclassification until 22.046%. While bagging for WHO-NCHS nutritional status can increase accurate classification from single data set 75.863% at 150 bootstrap replicated.   Keywords : Child nutritional status, Bagging, Ordinal logistics regression.